Journal of Urban Economics 59 (2006) 440–457www.elsevier.com/locate/jueZoning, TDRs and the density of development✩Virginia McConnella,b,∗, Margaret Wallsb, Elizabeth KopitscaUniversity of Maryland, Baltimore County, MDbResources for the Future, 1616 P Street, NW, Washington, DC 20036cUS EPA—National Center for Environmental Economics, 1200 Pennsylvania Avenue, NW, MC 1809T,Washington, DC 20460Received 16 May 2005; revised 12 December 2005Available online 3 February 2006AbstractThis paper explores the effect of low density zoning regulations and other factors on subdivision density.Using a unique dataset of new subdivisions built over a 34-year period in Calvert County, Maryland, weeconometrically estimate a density function using both OLS and censored regression. The variability indensity permitted by the county’s zoning and TDR rules over the sample period allows us to assess therelative importance of market factors and regulatory constraints on density. We use the censored model topredict what density patterns would have been without zoning.Published by Elsevier Inc.1. IntroductionMany communities on the urban fringe are implementing a range of policies to preservefarmland and open space, cluster residential development, and guide development to areas withexisting infrastructure. These efforts are an attempt to control overall growth and to counter atrend toward so-called “large lot” development so that growth that does take place results inless consumption of land (Heimlich and Anderson [14]). Planners have argued that policies tomanage density are the most important local policy focus for urban areas in the coming years(Danielson et al. [4]).✩The views expressed in this paper are those of the authors and do not necessarily represent those of the USEnvironmental Protection Agency. No official Agency endorsement should be inferred.*Corresponding author.E-mail addresses: [email protected] (V. McConnell), [email protected] (M. Walls), [email protected](E. Kopits).0094-1190/$ – see front matter Published by Elsevier Inc.doi:10.1016/j.jue.2005.12.005V. McConnell et al. / Journal of Urban Economics 59 (2006) 440–457 441Some researchers contend that large lot development and “sprawl” more generally are sim-ply the natural result of household preferences and market forces (Gordon and Richardson [12]).Glaeser and Kahn [11] argue that the widespread use of the car as a means of travel has made scat-tered, low-density development an inevitable market outcome.1Davis, Nelson, and Dueker [5]report results of a survey finding that 60% of people who move to so-called “ex-urban” locationsbeyond traditional suburbs move there to have large lots and a rural lifestyle.Other authors have argued that local government zoning rules in the form of large minimumlot-sizes may well be contributing to current patterns of low density development. Fischel [8]suggests that growth controls in the form of large lot zoning tend to result in “suburban sprawl.”There have been theoretical analyses showing the conditions under which this might be true(Moss [24], Pasha [26]), but there has been very little empirical work providing any evidence onthis issue.In this paper, drawing on a unique dataset, we are able to address two important questions.Do zoning regulations play a role in creating low density, land-intensive development, or is itprimarily market forces that generate such land-use patterns? The second is, if zoning limitsaccount for low-density development in at least some cases, how would development patterns bedifferent if there had been no such rules?We address these issues by analyzing the factors that explain subdivision density in a rapidlygrowing county on the fringes of the Washington, DC, metropolitan area, Calvert County, Mary-land. Economic variables that influence density are identified, including factors that affect thevalue and cost of additional development. Regulatory constraints and rules on building are in-cluded in the model. Calvert’s long-running transferable development rights (TDRs) programthat allows developers to increase density above base zoning limits in some areas by purchasingTDRs is also incorporated. We econometrically estimate the density function in both an OLSand a censored regression framework using detailed data on subdivisions built in Calvert Countyover a 34-year period. The variability in density permitted by the County’s zoning and TDR rulesover the sample period allows us to empirically assess the relative importance of market factorsand regulatory constraints on density. We then use the censored model to predict, in the caseswhere regulatory constraints have been binding, what density patterns would have been withoutthe zoning constraints.Although there is a large empirical literature looking at population and employment densities,this literature tends to focus on how density changes as distance to the central business districtof a metropolitan area increases and how zoning allocations—i.e., the proportion of land zonedresidential, commercial, industrial—affect overall density in a wide geographical area.2Amongthe few studies that analyze residential housing density at a smaller regional scale, Song andKnaap [29] use Census block data to examine how neighborhood density and other measuresof urban form have changed in the Portland, Oregon, metropolitan region. The authors find thatPortland’s strict development policies appear to be having some success in increasing single-family dwelling densities over time. Peiser [27] in an analysis of density patterns in three differentcommunities finds that discontinuous or infill development results in higher overall density in asuburban region than would occur if only continuous or sequential development was permitted.Thorsnes [31] focuses on the question of whether larger subdivisions, by being better able to1Many studies have emphasized the role played by declining transportation costs (Brueckner [1]); Glaeser and Kahn’s[11] particular point of emphasis is that the car has eliminated the scale economies that existed with older transportationtechnologies such as ports and railroad hubs.2McDonald [20] surveys this literature which draws from the theoretical base of Muth [25] and Mills [23].442 V. McConnell et al. / Journal of Urban Economics 59 (2006) 440–457internalize neighborhood externalities, lead to higher property values than smaller subdivisions,all else equal.